Diagnostic Performance

Abstract

The performance of a diagnostic examination¹ can be basically considered as its degree of accuracy, namely its ability to find the subjects affected with a given disease as positive and the subjects not affected with same disease as negative. The indices which in different ways measure this performance are defined measures of diagnostic performance and the studies aimed at measuring the diagnostic performance of an examination or, more often, at comparing the diagnostic performance of two or more examinations, are defined studies of diagnostic performance.

For the sake of clarity, we will avoid naming a radiologic examination as a test as much as possible. Although this term is entirely correct, we prefer to use the term exam or examination in order to avoid confusion with statistical tests. Exceptions will be pretest probability and post-test probability for a given disease which we will approach with Bayes’ theorem to represent the ability of each diagnostic test to increase or decrease the probability of a given disease in the subjects who underwent the examination with a positive or a negative result, respectively. Other rare exceptions will be evident from the context.